Data analysis and processing method and device based on artificial intelligence AI scalable model

Abstract

The present invention provides a data analysis and processing method and device based on an artificial intelligence (AI) scalable model, including: a server determines a corresponding first initial matching model according to a first attribute of a first user terminal; the first initial matching model generates a first dimension extension table of all matching dimensions according to the first attribute of the first user, and the server sends the first dimension extension table to the first user terminal; a second dimension extension table is obtained according to an adjusted first quantization standard and / or a new added quantization standard; the first initial matching model is adjusted to obtain an analysis and processing model, and the second attributes of other second user terminals are analyzed to obtain corresponding first scoring information; other second user terminals are the main body, and the first attributes are analyzed according to the corresponding analysis and processing model to obtain second scoring information, matching scoring information is obtained based on the first scoring information and the second scoring information, and the matching first user terminal and the second user terminal are pushed as analysis results based on the matching scoring information.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a data analysis and processing method and device based on an artificial intelligence (AI) scalable model. Background Art

[0002] As the market of social software grows, the accuracy requirements for matching multiple users in social software are getting higher and higher. In the existing technology, when recommending users, they are mostly divided by location, and other users' matching is randomly pushed, which makes it impossible to perform customized demand matching. In the matching process, one user is always matched as the subject, which means that the other subject may have a lower matching demand and cannot perform the final real match. Therefore, how to perform customized demand matching and how to perform forward and reverse fusion matching with multiple subjects as the target are urgent problems to be solved. Summary of the invention

[0003] The embodiments of the present invention provide a data analysis and processing method and device based on an artificial intelligence (AI) scalable model, which can customize the initial matching model according to user needs, so that the needs can be customized and matched. In addition, the present invention will first perform a preliminary comparison and a detailed comparison during matching. During the detailed comparison, forward and reverse fusion matching can be performed with multiple subjects as targets to obtain the final matching reference basis, so that the results obtained by the AI ​​scalable model provided by the present invention are easier for the subject to complete real matching offline, so that the matching degrees between each other are relatively the highest, thereby improving the matching efficiency.

[0004] A first aspect of an embodiment of the present invention provides a data analysis and processing method based on an artificial intelligence (AI) scalable model, comprising:

[0005] The server determines a corresponding first initial matching model according to the first attribute of the first user terminal, wherein the first initial matching model includes a plurality of first matching units, each of which has a first quantitative standard of a corresponding matching dimension;

[0006] The first initial matching model generates a first dimension extension table of all matching dimensions according to the first attribute of the first user, wherein the first dimension extension table has a first quantization standard for each matching dimension, and the server sends the first dimension extension table to the first user terminal;

[0007] If it is determined that the first user terminal adjusts at least one first quantization standard in the first dimension extension table and / or adds a new newly added quantization standard, a second dimension extension table is obtained according to the adjusted first quantization standard and / or the new newly added quantization standard;

[0008] The server adjusts the first initial matching model based on the second dimension extension table to obtain an analysis and processing model, and analyzes the second attributes of other second user terminals based on the analysis and processing model to obtain corresponding first scoring information;

[0009] The server again analyzes the first attribute based on the corresponding analysis and processing model with the other second user terminal as the main body to obtain second scoring information, obtains matching scoring information based on the first scoring information and the second scoring information, and pushes the matching first user terminal and second user terminal as analysis results based on the matching scoring information.

[0010] Optionally, in a possible implementation manner of the first aspect, the server determines a corresponding first initial matching model according to the first attribute of the first user terminal, where the first initial matching model includes a plurality of first matching units, each of the first matching units having a first quantitative standard of a corresponding matching dimension, including:

[0011] The server obtains a category parameter in the first attribute of the first user terminal and quantization parameters of different matching dimensions, and determines a plurality of reference parameter intervals of corresponding matching dimensions according to the category parameter;

[0012] Determine a calculation method of a quantization parameter and / or a preset adjustment parameter according to the type parameter, and calculate a target parameter according to the quantization parameter and the preset adjustment parameter;

[0013] After comparing the target parameter with a plurality of reference parameter intervals, a corresponding score value is added to each reference parameter interval to obtain a first quantitative standard having a corresponding matching dimension for each first matching unit.

[0014] Optionally, in a possible implementation manner of the first aspect, adding a corresponding score value to each benchmark parameter interval after comparing the target parameter with the multiple benchmark parameter intervals to obtain a first quantitative standard having a corresponding matching dimension for each first matching unit includes:

[0015] Determine the reference parameter interval in which the target parameter is located as the first reference parameter interval, determine the reference parameter interval in which the target parameter is not located as the second reference parameter interval, and calculate the distance value between the middle value of the second reference parameter interval and the target parameter;

[0016] sorting the second reference parameter interval in ascending order according to the distance value, and adding the first reference parameter interval to the head of the interval sorting sequence to obtain an interval sorting sequence;

[0017] A corresponding score value is added according to the position of each reference parameter interval in the interval sorting sequence, and reference parameter intervals at different positions have different score values.

[0018] Optionally, in a possible implementation manner of the first aspect, if it is determined that the first user terminal adjusts at least one first quantization standard in the first dimension extension table and / or adds a new newly added quantization standard, obtaining the second dimension extension table according to the adjusted first quantization standard and / or the new newly added quantization standard includes:

[0019] If it is determined that the first user terminal selects any matching dimension in the first dimension extension table, the slots corresponding to the score values ​​corresponding to the reference parameter intervals of the corresponding matching dimension are highlighted, so that the first user terminal adjusts the score value corresponding to the highlighted slot;

[0020] If it is determined that the first user terminal selects an extension column in the first dimension extension table, a preset extension category corresponding to the extension column is generated, and the newly added preset extension category and the score value corresponding to each quantitative interval or non-quantitative interval in the preset extension category are determined according to the interaction of the first user terminal;

[0021] After receiving the confirmation information, the first user terminal collects statistics of the adjusted first quantization standard and / or adds a new quantization standard to obtain a second dimension expansion table.

[0022] Optionally, in a possible implementation manner of the first aspect, the server adjusts the first initial matching model based on the second dimension extension table to obtain an analysis and processing model, and analyzes the second attributes of other second user terminals based on the analysis and processing model to obtain corresponding first scoring information, including:

[0023] The server compares the first dimension extension table with the second dimension extension table to obtain model adjustment data;

[0024] If it is determined that the model adjustment data includes an adjustment to the first quantization standard, adjusting the first matching unit and the first quantization standard of the first initial matching model based on the model adjustment data to obtain an analysis and processing model;

[0025] If it is determined that there is a newly added quantization standard in the model adjustment data, a new first matching unit and a first quantization standard are established for the first initial matching model based on the model adjustment data to obtain an analysis and processing model;

[0026] The matching information between the second attribute of the second user terminal and all corresponding matching dimensions in the analysis and processing model is obtained, and the matching information is input into the analysis and processing model to obtain corresponding first scoring information.

[0027] Optionally, in a possible implementation manner of the first aspect, if it is determined that the model adjustment data contains an adjustment to the first quantization standard, adjusting the first matching unit and the first quantization standard of the first initial matching model based on the model adjustment data to obtain an analysis and processing model includes:

[0028] If the first quantization standard is adjusted in the judgment model adjustment data, determining a corresponding first matching unit according to the adjusted matching dimension;

[0029] The score value of the adjusted reference parameter interval is replaced with the corresponding original score value in the original first quantization standard of the first matching unit to obtain an adjusted analysis and processing model.

[0030] Optionally, in a possible implementation manner of the first aspect, obtaining matching information between the second attribute of the second user terminal and all corresponding matching dimensions in the analysis and processing model, and inputting the matching information into the analysis and processing model to obtain corresponding first scoring information includes:

[0031] Obtaining matching information between the second attribute of the second user terminal and all corresponding matching dimensions in the analysis and processing model, wherein each second attribute has matching information uploaded in advance;

[0032] Input each matching information into the corresponding first matching unit in the analysis and processing model, and the first matching unit obtains a corresponding score value according to the reference parameter interval or non-quantization interval in which it is located;

[0033] Sum and average the score values ​​of all first matching units to obtain mean score information;

[0034] A matching coefficient is obtained according to the number of first matching units to which matching information has not been input and the number of first matching units to which matching information has been input, and the mean scoring information is adjusted according to the matching coefficient to obtain the first scoring information.

[0035] Optionally, in a possible implementation manner of the first aspect, obtaining a matching coefficient according to the number of first matching units to which matching information has not been input and the number of first matching units to which matching information has been input, and adjusting the mean scoring information according to the matching coefficient to obtain the first scoring information includes:

[0036] Dividing the number of first matching units to which matching information has not been input by the number of first matching units to which matching information has been input to obtain a proportionality coefficient, and weighting the proportionality coefficient to obtain a matching coefficient;

[0037] The first scoring information is obtained by subtracting the matching coefficient from the mean scoring information.

[0038] Optionally, in a possible implementation manner of the first aspect, the server again analyzes the first attribute according to the corresponding analysis and processing model with another second user terminal as the main body to obtain second scoring information, obtains matching scoring information based on the first scoring information and the second scoring information, and pushes the matching first user terminal and the second user terminal as analysis results based on the matching scoring information, including:

[0039] Sort the plurality of second user terminals in descending order based on the first scoring information to obtain a primary matching sequence, and select a first preset number of second user terminals in the user terminal matching sequence as the second user terminals to be recalculated;

[0040] The server analyzes the first attribute according to the corresponding analysis and processing model for the second user terminal to obtain second scoring information, and adds the first scoring information and the second scoring information to obtain the matching scoring information;

[0041] Based on the matching score information, the first preset number of second user terminals are sorted in descending order again to obtain a secondary matching sequence, a second preset number of second user terminals at the front of the secondary matching sequence are selected as second user terminals matching the first user terminal, and the analysis result is pushed to the first user terminal.

[0042] A second aspect of an embodiment of the present invention provides a data analysis and processing device based on an artificial intelligence (AI) scalable model, comprising:

[0043] A determination module, configured to enable the server to determine a corresponding first initial matching model according to a first attribute of the first user terminal, wherein the first initial matching model includes a plurality of first matching units, each of which has a first quantitative standard of a corresponding matching dimension;

[0044] A generating module, configured to enable the first initial matching model to generate a first dimension extension table of all matching dimensions according to the first attribute of the first user, wherein the first dimension extension table has a first quantization standard for each matching dimension, and the server sends the first dimension extension table to the first user terminal;

[0045] A judgment module, configured to obtain a second dimension extension table according to the adjusted first quantization standard and / or the new newly added quantization standard if it is judged that the first user terminal adjusts at least one first quantization standard in the first dimension extension table and / or adds a new newly added quantization standard;

[0046] an adjustment module, configured to enable the server to adjust the first initial matching model based on the second dimension extension table to obtain an analysis and processing model, and to analyze the second attributes of other second user terminals based on the analysis and processing model to obtain corresponding first scoring information;

[0047] The push module is used to enable the server to analyze the first attribute again with other second user terminals as the main body according to the corresponding analysis and processing model to obtain second scoring information, obtain matching scoring information based on the first scoring information and the second scoring information, and push the matching first user terminal and the second user terminal as analysis results based on the matching scoring information.

[0048] According to a third aspect of an embodiment of the present invention, a storage medium is provided, in which a computer program is stored. When the computer program is executed by a processor, it is used to implement the method of the first aspect of the present invention and various possible designs of the first aspect.

[0049] The present invention provides a data analysis and processing method and device based on an artificial intelligence (AI) scalable model, which can customize the initial matching model according to user needs, so that the needs can be customized and matched. In addition, the present invention will first perform a preliminary comparison and a detailed comparison during matching. During the detailed comparison, forward and reverse fusion matching can be performed with multiple subjects as targets to obtain the final matching reference basis, so that the results obtained by the AI ​​scalable model provided by the present invention are easier for the subject to complete real matching offline, so that the mutual matching degrees are relatively the highest, thereby improving the matching efficiency.

[0050] The technical solution provided by the present invention will obtain a corresponding second dimension extension table based on the user's adjustment or new addition of the matching dimension in the first dimension extension table, and after obtaining the second dimension extension table, the present invention will adjust the original AI model based on the model adjustment data of the first dimension extension table to obtain an adjusted analysis and processing model. This method enables the present invention to dynamically customize and adjust the AI ​​model in an interactive manner, so that the subsequently calculated values ​​reflecting the matching Chengdu can be in line with the user's needs.

[0051] When calculating the second user terminal corresponding to the first user terminal, the AI ​​model in the present invention will first determine a certain number of matching second user terminals based on the first user terminal, and then perform reverse calculation based on the determined second user terminals to obtain the matching score of the first user terminal, and then perform a comprehensive calculation based on the two scores to obtain the final matching score. This method enables the present invention to perform preliminary screening and fine screening on the corresponding second user terminals. The preliminary screening and fine screening methods are different. While ensuring that the first user terminal and the second user terminal have a high degree of matching, the data processing amount is reduced, thereby improving the processing efficiency of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a flow chart of a data analysis and processing method based on an artificial intelligence (AI) scalable model;

[0053] Figure 2 This is a structural diagram of a data analysis and processing device based on an artificial intelligence (AI) scalable model. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in sequences other than those illustrated or described herein.

[0056] It should be understood that in various embodiments of the present invention, the size of the sequence number of each process does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0057] It should be understood that in the present invention, "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or apparatus comprising a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products or apparatuses.

[0058] It should be understood that in the present invention, "plurality" refers to two or more than two. "And / or" is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "Contains A, B and C", "Contains A, B, C" means that A, B, and C are all included, "Contains A, B or C" means that one of A, B, and C is included, and "Contains A, B and / or C" means that any one, any two, or any three of A, B, and C are included.

[0059] It should be understood that in the present invention, "B corresponding to A", "B corresponding to A", "A corresponds to B" or "B corresponds to A" means that B is associated with A and B can be determined based on A. Determining B based on A does not mean determining B based only on A, but B can also be determined based on A and / or other information. A and B match when the similarity between A and B is greater than or equal to a preset threshold.

[0060] Depending on the context, "if" as used herein may be interpreted as "when" or "when" or "in response to determining" or "in response to detecting."

[0061] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0062] The present invention provides a data analysis and processing method based on an artificial intelligence (AI) scalable model, such as Figure 1 As shown, including:

[0063] Step S110, the server determines the corresponding first initial matching model according to the first attribute of the first user terminal, and the first initial matching model includes multiple first matching units, and each first matching unit has a first quantitative standard of the corresponding matching dimension. The server in the present invention will determine the corresponding first initial matching model according to the first attribute of the first user terminal, and the first attribute at least includes gender. The first initial matching model corresponding to the present invention when matching male users or female users is different. Different first initial matching models include multiple first matching units of the same type, but the first quantitative standards of the corresponding matching dimensions are generally different. For example, the matching dimension of the first matching unit in men can be height. Generally speaking, girls are shorter than boys, which makes the two more matched, so the initial first matching unit will determine the first quantitative standard of the corresponding matching dimension according to the height of boys.

[0064] In a possible implementation of the technical solution provided by the present invention, step S110 includes:

[0065] The server obtains the category parameter in the first attribute of the first user terminal and the quantization parameters of different matching dimensions, and determines multiple benchmark parameter intervals of the corresponding matching dimensions according to the category parameter. In the present invention, the server will actively obtain the category parameter in the first attribute of the first user terminal, such as height, weight, age, income, etc., and the quantization parameter is the numerical value corresponding to the corresponding height, weight, age, and income. The present invention will determine multiple benchmark parameter intervals of the corresponding matching dimensions according to the category parameter, such as height 151cm to 155cm, 156cm to 160cm, 161cm to 165cm, 166cm to 170cm, etc.

[0066] Determine the calculation method of the quantization parameter and / or the preset adjustment parameter according to the type parameter, and calculate the target parameter according to the quantization parameter and the preset adjustment parameter. The present invention will determine the calculation method of the quantization parameter according to the type parameter. If the type parameter is a non-numerical parameter, such as an appearance evaluation, such as high, medium, low, etc. in appearance evaluation, the present invention will perform corresponding quantization. Taking the type parameter as the height distance, the present invention will determine the preset adjustment parameter corresponding to the height distance. The preset adjustment parameter can be -10cm, that is, the girl is 10 centimeters shorter than the boy as the system default best matching scheme for men and women. The present invention will be based on the quantization parameter 175cm (for example, the boy is 175cm tall), then the target parameter at this time is 175cm-10cm=165cm, and the calculation method is a subtraction calculation.

[0067] After comparing the target parameter with multiple reference parameter intervals, a corresponding score value is added to each reference parameter interval to obtain a first quantitative standard with a corresponding matching dimension for each first matching unit. The present invention compares the target parameter with multiple reference parameter intervals. At this time, 165cm is within the reference parameter interval of 161cm to 165cm, and the corresponding reference parameter interval of 161cm to 165cm will match the highest score value.

[0068] In a possible implementation of the technical solution provided by the present invention, after comparing the target parameter with a plurality of reference parameter intervals, adding a corresponding score value to each reference parameter interval to obtain a first quantitative standard having a corresponding matching dimension for each first matching unit includes:

[0069] Determine the reference parameter interval where the target parameter is located as the first reference parameter interval, and the reference parameter interval where the target parameter is not located as the second reference parameter interval, and calculate the distance value between the interval intermediate value of the second reference parameter interval and the target parameter. The present invention uses the reference parameter interval of 161cm to 165cm as the first reference parameter interval, and the reference parameter intervals of 151cm to 155cm, 156cm to 160cm, 166cm to 170cm as the second reference parameter interval, and the present invention calculates the distance value between the interval intermediate value of the second reference parameter interval and the target parameter, for example, the distance value between the reference parameter interval of 166cm to 170cm and the target parameter is 1, and the distance value between the reference parameter interval of 156cm to 160cm and the target parameter is 5.

[0070] The second reference parameter interval is sorted in ascending order according to the distance value, and the first reference parameter interval is added to the head of the interval sorting sequence to obtain the interval sorting sequence. In this way, the present invention will make the reference parameter interval closer to 165cm be ranked in the front to obtain the corresponding interval sorting sequence.

[0071] According to the position of each reference parameter interval in the interval sorting sequence, a corresponding score value is added, and reference parameter intervals at different positions have different score values. The present invention pre-sets different score values ​​for different positions. For example, the score value of the first position in the interval sorting sequence is 10, and the score value of the second position is 8. By reducing in this way, reference parameter intervals at different positions have different score values. The score value can reflect the matching degree of other users matched with the user in height dimension and other dimensions.

[0072] Step S120: The first initial matching model generates a first dimension extension table of all matching dimensions according to the first attribute of the first user, and the first dimension extension table has a first quantitative standard for each matching dimension, and the server sends the first dimension extension table to the first user terminal. The first initial matching model will generate a first dimension extension table of all matching dimensions by counting the first attribute of the first user, and display the matching rules and regulations to the user through the first dimension extension table, so that the corresponding interactive adjustments can be made according to the actual needs of the user in the future, and the first dimension extension table has a first quantitative standard for each matching dimension, and the server sends the first dimension extension table to the first user terminal.

[0073] Step S130: If it is determined that the first user terminal adjusts at least one first quantitative standard in the first dimension extension table and / or adds a new quantitative standard, then the second dimension extension table is obtained according to the adjusted first quantitative standard and / or the new quantitative standard. When the user needs to modify or add adjustments to the first dimension extension table, the present invention will obtain the second dimension extension table according to the adjusted first quantitative standard and / or the new quantitative standard. The second dimension extension table is the matching requirement after the user performs specialized customization. For example, if the user likes people who are 5 cm shorter than him, then the corresponding first benchmark parameter area is 166 cm to 170 cm, and the user can adjust the first quantitative standard for the height dimension.

[0074] In a possible implementation of the technical solution provided by the present invention, step S130 includes:

[0075] If it is determined that the first user terminal has selected any matching dimension in the first dimension expansion table, the slots corresponding to the score values ​​corresponding to the benchmark parameter intervals of the corresponding matching dimension are highlighted, so that the first user terminal can adjust the score values ​​corresponding to the highlighted slots. When adjusting a matching dimension relative to a certain matching dimension, the user will select the corresponding matching dimension, and highlight the slots corresponding to the score values ​​corresponding to the benchmark parameter intervals of the corresponding matching dimension. At this time, the user can adjust the score values ​​corresponding to the highlighted slots, for example, adjusting the values ​​of 166cm to 170cm to 10, adjusting the values ​​of 161cm to 165cm to 6, and so on.

[0076] If it is determined that the first user terminal selects an extension column in the first dimension extension table, a preset extension category corresponding to the extension column is generated, and the newly added preset extension category and the score value corresponding to each quantitative interval or non-quantitative interval in the preset extension category are determined according to the interaction of the first user terminal. The present invention will pre-set extension columns, and the extension columns can be expanded and added by users, such as income, type of work, etc. The present invention will be based on the preset extension categories newly added by the user and the score values ​​corresponding to each quantitative interval or non-quantitative interval in the preset extension category. For example, if the newly added preset extension category is work, the present invention will display the preset types of work, such as civil servants, state-owned enterprise employees, lawyers, etc. At this time, users can add corresponding score values ​​to the types of work in different non-quantitative intervals.

[0077] After receiving the confirmation information, the first user terminal statistically adjusts the adjusted first quantitative standard and / or adds a new quantitative standard to obtain a second dimension expansion table. After receiving the confirmation information, the first user terminal proves that the user has modified and agreed to the automatically generated matching dimension, and no new quantitative standard will be added. At this time, the second dimension expansion table is the user's optimal matching strategy and matching rule.

[0078] Step S140: The server adjusts the first initial matching model based on the second dimension extension table to obtain an analysis and processing model, and analyzes the second attributes of other second user terminals based on the analysis and processing model to obtain corresponding first scoring information. After receiving the second dimension extension table, the server adjusts the first initial matching model to obtain an analysis and processing model, which is an analysis and processing model corresponding to the user's own matching needs after customization. The present invention analyzes the second attributes of other second user terminals based on the analysis and processing model to obtain corresponding first scoring information.

[0079] In a possible implementation of the technical solution provided by the present invention, step S140 includes:

[0080] The server compares the first dimension extension table with the second dimension extension table to obtain model adjustment data. The present invention compares the first dimension extension table with the second dimension extension table to obtain adjustment data for the first initial matching model, so that the present invention can quickly adjust the first initial matching model based on the adjustment data.

[0081] If it is determined that there is an adjustment to the first quantitative standard in the model adjustment data, the first matching unit and the first quantitative standard of the first initial matching model are adjusted based on the model adjustment data to obtain an analysis and processing model. At this time, the user adjusts the first quantitative standard of a matching dimension in the first initial matching model, and the present invention adjusts the first matching unit and the first quantitative standard of the first initial matching model according to the model adjustment data to obtain an analysis and processing model.

[0082] In a possible implementation of the technical solution provided by the present invention, if it is determined that the model adjustment data contains an adjustment to the first quantitative standard, then based on the model adjustment data, the first matching unit and the first quantitative standard of the first initial matching model are adjusted to obtain an analysis and processing model, including:

[0083] If the model adjustment data determines that the first quantization standard is adjusted, the corresponding first matching unit is determined according to the adjusted matching dimension. At this time, the present invention first determines the first matching unit corresponding to the adjusted matching dimension, and then adjusts the subsequent information accordingly.

[0084] The score value of the adjusted reference parameter interval is replaced with the corresponding original score value in the original first quantitative standard of the first matching unit to obtain an adjusted analysis and processing model. The present invention replaces the score value of the adjusted reference parameter interval with the corresponding original score value in the original first quantitative standard of the first matching unit, for example, the value of 166cm to 170cm is adjusted to 10, the value of 161cm to 165cm is adjusted to 6, and so on.

[0085] If it is determined that there is a newly added quantitative standard in the model adjustment data, a new first matching unit and a first quantitative standard are established for the first initial matching model based on the model adjustment data to obtain an analysis and processing model. In actual application scenarios, users may add new matching dimensions, such as the above-mentioned income, work and other categories and the first quantitative standard corresponding to each category. The new first matching unit and the first quantitative standard can be established by the first user completely according to his own needs.

[0086] Obtain the matching information between the second attribute of the second user terminal and all corresponding matching dimensions in the analysis and processing model, and input the matching information into the analysis and processing model to obtain the corresponding first scoring information. The present invention selects the second attribute of the second user terminal according to the adjusted analysis and processing model. The second attribute of the second user terminal is uploaded by the second user, and the second attribute can be multiple. The present invention obtains the matching information between the second attribute and all corresponding matching dimensions in the analysis and processing model, that is, to obtain the matching information about the second user that has scoring significance required by the first user. Finally, the matching information is calculated and processed based on the analysis and processing model to obtain the first scoring information.

[0087] In a possible implementation of the technical solution provided by the present invention, the step of obtaining matching information between the second attribute of the second user terminal and all corresponding matching dimensions in the analysis and processing model, and inputting the matching information into the analysis and processing model to obtain corresponding first scoring information includes:

[0088] The matching information of the second attribute of the second user terminal and all corresponding matching dimensions in the analysis and processing model is obtained, and each second attribute has matching information uploaded in advance. The present invention obtains the matching information of the second attribute and all corresponding matching dimensions in the analysis and processing model, and the information of the unmatched matching dimensions is no longer obtained, so that the present invention only obtains the matching information with the first user as the main body and having a scoring effect.

[0089] Each matching information is input into the corresponding first matching unit in the analysis and processing model, and the first matching unit obtains a corresponding score value according to the reference parameter interval or non-quantization interval in which it is located. The present invention inputs each matching information into the corresponding first matching unit in the analysis and processing model, and obtains a corresponding score value according to the reference parameter interval or non-quantization interval in which the matching information is located.

[0090] The score values ​​of all first matching units are summed and averaged to obtain the mean score information. In the above manner, the present invention can calculate the mean of all matching dimensions to obtain the final mean score information, which is the score obtained by the dimension that the first user is relatively more concerned about, and has a higher matching degree with the first user. The matching dimensions corresponding to different first users may be different.

[0091] A matching coefficient is obtained according to the number of first matching units that have not been input with matching information and the number of first matching units that have been input with matching information, and the first scoring information is obtained by adjusting the mean scoring information according to the matching coefficient. In some scenarios, the matching information of the user is not completely matched with the first matching unit of the analysis and processing model, and the first matching unit has a matching dimension but the user does not have the corresponding matching information. Therefore, the present invention obtains a matching coefficient according to the number of first matching units that have not been input with matching information and the number of first matching units that have been input with matching information, and finally adjusts the mean scoring information in combination with the matching coefficient to obtain the first scoring information.

[0092] In a possible implementation manner of the technical solution provided by the present invention, obtaining a matching coefficient according to the number of first matching units to which matching information has not been input and the number of first matching units to which matching information has been input, and adjusting the mean scoring information according to the matching coefficient to obtain the first scoring information includes:

[0093] The number of first matching units that have not been input with matching information is divided by the number of first matching units that have been input with matching information to obtain a proportional coefficient, and the proportional coefficient is weighted to obtain a matching coefficient. The present invention divides the number of first matching units that have not been input with matching information by the number of first matching units that have been input with matching information. The larger the proportional coefficient, the more matching information is missing. The larger the proportional coefficient is at this time. If the number of first matching units that have not been input with matching information is 0, the calculated proportional coefficient is also 0. The present invention weights the proportional coefficient to obtain a matching coefficient. The weighting method may be multiplying by a preset value, such as 3, 5, 10, etc. The preset value for weighting may be pre-configured by the administrator according to the number of first matching units in the first initial matching model.

[0094] The first scoring information is obtained by subtracting the matching coefficient from the mean scoring information. This method makes it possible for the relative score of the second user with fewer types of corresponding matching information to be lower, and thus when matching the first user with the second user, the present invention will make a reference based on the number of matched matching dimensions, so that the obtained first scoring information is more objective, has more reference dimensions, and has a higher matching property.

[0095] Step S150, the server again analyzes the first attribute according to the corresponding analysis and processing model with other second user terminals as the main body to obtain second scoring information, obtains matching scoring information based on the first scoring information and the second scoring information, and pushes the matching first user terminal and the second user terminal as analysis results based on the matching scoring information. In the technical solution provided by the present invention, after determining the score of the second user terminal by determining the score of the second user terminal with the first user terminal, the present invention analyzes the first attribute according to the corresponding analysis and processing model with other second user terminals as the main body to obtain second scoring information. In actual matching scenarios, there is an essential difference between matching people and matching people and objects. When matching people, the first user needs to feel that the second user matches him, and the second user also needs to feel that the first user matches him. This scenario is a real match. If only the first user feels that the second user matches, but the second user does not feel that the first user matches, then the first user and the second user are not considered a match. Therefore, the server in the present invention will again analyze the first attribute according to the corresponding analysis and processing model with other second user terminals as the main body to obtain the second scoring information, that is, the analysis and processing model corresponding to the second user terminal will analyze the first attribute to obtain the second scoring information, and the way to obtain the analysis and processing model corresponding to the second user terminal can be the same as or corresponding to the way to obtain the analysis and processing model corresponding to the first user terminal. The present invention will obtain the final matching score information based on the pairwise scores between the first user terminal and each second user terminal, and finally push the matching first user terminal and second user terminal as the analysis result based on the matching score information.

[0096] In a possible implementation of the technical solution provided by the present invention, step S140 includes:

[0097] Based on the first scoring information, multiple second user terminals are sorted in descending order to obtain a matching sequence, and a first preset number of second user terminals in the front of the user terminal matching sequence are selected as the second user terminals to be calculated again. The present invention will sort multiple second user terminals in descending order according to the first scoring information to obtain a matching sequence, and select a certain number of second user terminals. For example, the first preset number can be 100. This method makes the subsequent second user terminals quantitative when calculating the second scoring information of the first user terminal, thereby ensuring that the calculation amount of the server is greatly reduced and the calculation efficiency is improved. The present invention first selects a certain number of second user terminals that need to be calculated again according to the first scoring information.

[0098] The server recalculates the second user end as the main body and analyzes the first attribute according to the corresponding analysis and processing model to obtain the second scoring information, and adds the first scoring information and the second scoring information to obtain the matching scoring information. The server in the present invention recalculates the second user end as the main body and analyzes the first attribute according to the corresponding analysis and processing model to obtain the second scoring information. The way it calculates the second scoring information can be the same as the way the first user end calculates the first scoring information. The present invention adds the first scoring information and the second scoring information to obtain the matching scoring information. When adding and averaging, the weights of the first scoring information and the second scoring information can be different, and the weight of the first scoring information is preferably greater than the weight of the second scoring information.

[0099] Based on the matching score information, the first preset number of second user terminals are sorted in descending order again to obtain a secondary matching sequence, the second preset number of second user terminals in the front of the secondary matching sequence are selected as the second user terminals that match the first user terminal, and the analysis results are pushed to the first user terminal. The present invention will sort the first preset number of second user terminals in descending order again according to the matching score information to obtain a secondary matching sequence. At this time, the secondary matching sequence is the secondary matching sequence after the first user performs a forward match on the second user and the second user performs a reverse match on the first user. The present invention will select the second preset number of second user terminals in the front of the secondary matching sequence as the second user terminals that match the first user terminal. The second preset number is preferably less than the first preset number. The present invention will select the second preset number of second user terminals in the front of the secondary matching sequence as the second user terminals that match the first user terminal, and push the analysis results to the first user terminal, so that the first user terminal can understand the second user terminals that match it in both directions, and perform subsequent operations such as adding friends. The present invention will no longer explain and limit the subsequent steps.

[0100] In order to implement the data analysis and processing method based on the artificial intelligence AI scalable model provided by the present invention, the present invention also provides a data analysis and processing device based on the artificial intelligence AI scalable model, such as Figure 2 As shown, including:

[0101] A determination module, configured to enable the server to determine a corresponding first initial matching model according to a first attribute of the first user terminal, wherein the first initial matching model includes a plurality of first matching units, each of which has a first quantitative standard of a corresponding matching dimension;

[0102] A generating module, configured to enable the first initial matching model to generate a first dimension extension table of all matching dimensions according to the first attribute of the first user, wherein the first dimension extension table has a first quantization standard for each matching dimension, and the server sends the first dimension extension table to the first user terminal;

[0103] A judgment module, configured to obtain a second dimension extension table according to the adjusted first quantization standard and / or the new newly added quantization standard if it is judged that the first user terminal adjusts at least one first quantization standard in the first dimension extension table and / or adds a new newly added quantization standard;

[0104] an adjustment module, configured to enable the server to adjust the first initial matching model based on the second dimension extension table to obtain an analysis and processing model, and to analyze the second attributes of other second user terminals based on the analysis and processing model to obtain corresponding first scoring information;

[0105] The push module is used to enable the server to analyze the first attribute again with other second user terminals as the main body according to the corresponding analysis and processing model to obtain second scoring information, obtain matching scoring information based on the first scoring information and the second scoring information, and push the matching first user terminal and the second user terminal as analysis results based on the matching scoring information.

[0106] The present invention also provides a storage medium, in which a computer program is stored. When the computer program is executed by a processor, it is used to implement the methods provided by the various embodiments described above.

[0107] Among them, the storage medium can be a computer storage medium or a communication medium. The communication medium includes any medium that facilitates the transmission of a computer program from one place to another. The computer storage medium can be any available medium that can be accessed by a general-purpose or special-purpose computer. For example, the storage medium is coupled to the processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an application-specific integrated circuit (Application Specific Integrated Circuits, referred to as ASIC). In addition, the ASIC can be located in a user device. Of course, the processor and the storage medium can also exist in a communication device as discrete components. The storage medium can be a read-only memory (ROM), a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

[0108] The present invention also provides a program product, which includes an execution instruction, which is stored in a storage medium. At least one processor of a device can read the execution instruction from the storage medium, and at least one processor executes the execution instruction so that the device implements the methods provided in the above various embodiments.

[0109] In the above-mentioned terminal or server embodiments, it should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the present invention may be directly implemented as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.

[0110] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A data analysis and processing method based on an artificial intelligence (AI) scalable model, characterized in that: include: The server determines a corresponding first initial matching model according to the first attribute of the first user terminal, wherein the first initial matching model includes a plurality of first matching units, each of which has a first quantitative standard of a corresponding matching dimension; The first initial matching model generates a first dimension extension table of all matching dimensions according to the first attribute of the first user, wherein the first dimension extension table has a first quantization standard for each matching dimension, and the server sends the first dimension extension table to the first user terminal; If it is determined that the first user terminal adjusts at least one first quantization standard in the first dimension extension table and / or adds a new newly added quantization standard, a second dimension extension table is obtained according to the adjusted first quantization standard and / or the new newly added quantization standard; The server adjusts the first initial matching model based on the second dimension extension table to obtain an analysis and processing model, and analyzes the second attributes of other second user terminals based on the analysis and processing model to obtain corresponding first scoring information; The server again analyzes the first attribute based on the other second user terminal according to the corresponding analysis and processing model to obtain second scoring information, obtains matching scoring information based on the first scoring information and the second scoring information, and pushes the matching first user terminal and the second user terminal as analysis results based on the matching scoring information; The server determines a corresponding first initial matching model according to a first attribute of the first user terminal, wherein the first initial matching model includes a plurality of first matching units, each of which has a first quantitative standard of a corresponding matching dimension, including: The server obtains a category parameter in the first attribute of the first user terminal and quantization parameters of different matching dimensions, and determines a plurality of reference parameter intervals of corresponding matching dimensions according to the category parameter; Determine a calculation method of a quantization parameter and / or a preset adjustment parameter according to the type parameter, and calculate a target parameter according to the quantization parameter and the preset adjustment parameter; After comparing the target parameter with a plurality of reference parameter intervals, a corresponding score value is added to each reference parameter interval to obtain a first quantitative standard having a corresponding matching dimension for each first matching unit; The step of comparing the target parameter with a plurality of reference parameter intervals and adding a corresponding score value to each reference parameter interval to obtain a first quantitative standard having a corresponding matching dimension for each first matching unit includes: Determine the reference parameter interval in which the target parameter is located as the first reference parameter interval, determine the reference parameter interval in which the target parameter is not located as the second reference parameter interval, and calculate the distance value between the middle value of the second reference parameter interval and the target parameter; sorting the second reference parameter interval in ascending order according to the distance value, and adding the first reference parameter interval to the head of the interval sorting sequence to obtain an interval sorting sequence; Adding a corresponding score value according to the position of each reference parameter interval in the interval sorting sequence, where reference parameter intervals at different positions have different score values; If it is determined that the first user terminal adjusts at least one first quantization standard in the first dimension extension table and / or adds a new quantization standard, obtaining the second dimension extension table according to the adjusted first quantization standard and / or the new quantization standard includes: If it is determined that the first user terminal selects any matching dimension in the first dimension extension table, the slots corresponding to the score values ​​corresponding to the reference parameter intervals of the corresponding matching dimension are highlighted, so that the first user terminal adjusts the score value corresponding to the highlighted slot; If it is determined that the first user terminal selects an extension column in the first dimension extension table, a preset extension category corresponding to the extension column is generated, and the newly added preset extension category and the score value corresponding to each quantitative interval or non-quantitative interval in the preset extension category are determined according to the interaction of the first user terminal; After receiving the confirmation information, the first user terminal counts the adjusted first quantization standard and / or adds a new newly added quantization standard to obtain a second dimension expansion table; The server adjusts the first initial matching model based on the second dimension extension table to obtain an analysis and processing model, and analyzes the second attributes of other second user terminals based on the analysis and processing model to obtain corresponding first scoring information, including: The server compares the first dimension extension table with the second dimension extension table to obtain model adjustment data; If it is determined that the model adjustment data includes an adjustment to the first quantization standard, adjusting the first matching unit and the first quantization standard of the first initial matching model based on the model adjustment data to obtain an analysis and processing model; If it is determined that there is a newly added quantization standard in the model adjustment data, a new first matching unit and a first quantization standard are established for the first initial matching model based on the model adjustment data to obtain an analysis and processing model; The matching information between the second attribute of the second user terminal and all corresponding matching dimensions in the analysis and processing model is obtained, and the matching information is input into the analysis and processing model to obtain corresponding first scoring information.

2. The data analysis and processing method based on the artificial intelligence AI scalable model according to claim 1 is characterized in that: If it is determined that the model adjustment data contains an adjustment to the first quantization standard, then adjusting the first matching unit and the first quantization standard of the first initial matching model based on the model adjustment data to obtain an analysis and processing model includes: If the first quantization standard is adjusted in the judgment model adjustment data, determining a corresponding first matching unit according to the adjusted matching dimension; The score value of the adjusted reference parameter interval is replaced with the corresponding original score value in the original first quantization standard of the first matching unit to obtain an adjusted analysis and processing model.

3. The data analysis and processing method based on the artificial intelligence AI scalable model according to claim 1 is characterized in that: The obtaining of matching information between the second attribute of the second user terminal and all corresponding matching dimensions in the analysis and processing model, and inputting the matching information into the analysis and processing model to obtain corresponding first scoring information, includes: Obtaining matching information between the second attribute of the second user terminal and all corresponding matching dimensions in the analysis and processing model, wherein each second attribute has matching information uploaded in advance; Input each matching information into the corresponding first matching unit in the analysis and processing model, and the first matching unit obtains a corresponding score value according to the reference parameter interval or non-quantization interval in which it is located; Sum and average the score values ​​of all first matching units to obtain mean score information; A matching coefficient is obtained according to the number of first matching units to which matching information has not been input and the number of first matching units to which matching information has been input, and the mean scoring information is adjusted according to the matching coefficient to obtain the first scoring information.

4. The data analysis and processing method based on the artificial intelligence AI scalable model according to claim 3 is characterized in that: The obtaining of a matching coefficient according to the number of first matching units to which matching information has not been input and the number of first matching units to which matching information has been input, and adjusting the mean scoring information according to the matching coefficient to obtain the first scoring information includes: Dividing the number of first matching units to which matching information has not been input by the number of first matching units to which matching information has been input to obtain a proportionality coefficient, and weighting the proportionality coefficient to obtain a matching coefficient; The first scoring information is obtained by subtracting the matching coefficient from the mean scoring information.

5. The data analysis and processing method based on the artificial intelligence AI scalable model according to claim 4 is characterized in that: The server again analyzes the first attribute based on the other second user terminal according to the corresponding analysis and processing model to obtain second scoring information, obtains matching scoring information based on the first scoring information and the second scoring information, and pushes the matching first user terminal and the second user terminal as analysis results based on the matching scoring information, including: Sort the plurality of second user terminals in descending order based on the first scoring information to obtain a primary matching sequence, and select a first preset number of second user terminals in the user terminal matching sequence as the second user terminals to be recalculated; The server analyzes the first attribute according to the corresponding analysis and processing model for the second user terminal to obtain second scoring information, and adds the first scoring information and the second scoring information to obtain the average value to obtain matching scoring information; Based on the matching score information, the first preset number of second user terminals are sorted in descending order again to obtain a secondary matching sequence, a second preset number of second user terminals at the front of the secondary matching sequence are selected as second user terminals matching the first user terminal, and the analysis result is pushed to the first user terminal.

6. A data analysis and processing device based on an artificial intelligence (AI) scalable model, characterized in that: include: A determination module, configured to enable the server to determine a corresponding first initial matching model according to a first attribute of the first user terminal, wherein the first initial matching model includes a plurality of first matching units, each of which has a first quantitative standard of a corresponding matching dimension; A generating module, configured to enable the first initial matching model to generate a first dimension extension table of all matching dimensions according to the first attribute of the first user, wherein the first dimension extension table has a first quantization standard for each matching dimension, and the server sends the first dimension extension table to the first user terminal; A judgment module, configured to obtain a second dimension extension table according to the adjusted first quantization standard and / or the new newly added quantization standard if it is judged that the first user terminal adjusts at least one first quantization standard in the first dimension extension table and / or adds a new newly added quantization standard; an adjustment module, configured to enable the server to adjust the first initial matching model based on the second dimension extension table to obtain an analysis and processing model, and to analyze the second attributes of other second user terminals based on the analysis and processing model to obtain corresponding first scoring information; a push module, configured to enable the server to analyze the first attribute again with another second user terminal as the main body according to the corresponding analysis and processing model to obtain second scoring information, obtain matching scoring information based on the first scoring information and the second scoring information, and push the matching first user terminal and the second user terminal as analysis results based on the matching scoring information; The server determines a corresponding first initial matching model according to a first attribute of the first user terminal, wherein the first initial matching model includes a plurality of first matching units, each of which has a first quantitative standard of a corresponding matching dimension, including: The server obtains a category parameter in the first attribute of the first user terminal and quantization parameters of different matching dimensions, and determines a plurality of reference parameter intervals of corresponding matching dimensions according to the category parameter; Determine a calculation method of a quantization parameter and / or a preset adjustment parameter according to the type parameter, and calculate a target parameter according to the quantization parameter and the preset adjustment parameter; After comparing the target parameter with a plurality of reference parameter intervals, a corresponding score value is added to each reference parameter interval to obtain a first quantitative standard having a corresponding matching dimension for each first matching unit; The step of comparing the target parameter with a plurality of reference parameter intervals and adding a corresponding score value to each reference parameter interval to obtain a first quantitative standard having a corresponding matching dimension for each first matching unit includes: Determine the reference parameter interval in which the target parameter is located as the first reference parameter interval, determine the reference parameter interval in which the target parameter is not located as the second reference parameter interval, and calculate the distance value between the middle value of the second reference parameter interval and the target parameter; sorting the second reference parameter interval in ascending order according to the distance value, and adding the first reference parameter interval to the head of the interval sorting sequence to obtain an interval sorting sequence; Adding a corresponding score value according to the position of each reference parameter interval in the interval sorting sequence, where reference parameter intervals at different positions have different score values; If it is determined that the first user terminal adjusts at least one first quantization standard in the first dimension extension table and / or adds a new quantization standard, obtaining the second dimension extension table according to the adjusted first quantization standard and / or the new quantization standard includes: If it is determined that the first user terminal selects any matching dimension in the first dimension extension table, the slots corresponding to the score values ​​corresponding to the reference parameter intervals of the corresponding matching dimension are highlighted, so that the first user terminal adjusts the score value corresponding to the highlighted slot; If it is determined that the first user terminal selects an extension column in the first dimension extension table, a preset extension category corresponding to the extension column is generated, and the newly added preset extension category and the score value corresponding to each quantitative interval or non-quantitative interval in the preset extension category are determined according to the interaction of the first user terminal; After receiving the confirmation information, the first user terminal counts the adjusted first quantization standard and / or adds a new newly added quantization standard to obtain a second dimension expansion table; The server adjusts the first initial matching model based on the second dimension extension table to obtain an analysis and processing model, and analyzes the second attributes of other second user terminals based on the analysis and processing model to obtain corresponding first scoring information, including: The server compares the first dimension extension table with the second dimension extension table to obtain model adjustment data; If it is determined that the model adjustment data includes an adjustment to the first quantization standard, adjusting the first matching unit and the first quantization standard of the first initial matching model based on the model adjustment data to obtain an analysis and processing model; If it is determined that there is a newly added quantization standard in the model adjustment data, a new first matching unit and a first quantization standard are established for the first initial matching model based on the model adjustment data to obtain an analysis and processing model; The matching information between the second attribute of the second user terminal and all corresponding matching dimensions in the analysis and processing model is obtained, and the matching information is input into the analysis and processing model to obtain corresponding first scoring information.

Images